A concise brief for tomorrow’s meeting covering our libp2p-based decentralized networking projects, current infrastructure, and how Akamai Cloud fits into the multi-cloud strategy powering the Cisco AGNTCY initiative.
Why we are meeting and what we hope to accomplish.
libp2p/Protocol Labs runs AWS-based infrastructure today and is evaluating Akamai Cloud as part of a broader multi-cloud strategy for decentralized AI and IoT workloads.
Identify how Akamai Cloud can support py-ipfs-lite and Universal Connectivity scaling, and define a practical Proof of Concept aligned with Cisco’s AGNTCY initiative.
A tailored Akamai architecture recommendation, usage-credit plan, and clear next steps for a multi-cloud deployment with observability and cost governance.
The two technical workstreams we plan to scale with Akamai Cloud.
A lightweight IPFS implementation in Python. It provides content-addressed storage and retrieval primitives for applications that need IPFS compatibility without running a full Kubo node.
A libp2p-based connectivity initiative that makes peer-to-peer networking reliable across transports, NATs, and runtime environments. It is a core enabler for distributed applications.
Both projects are aligned with Cisco’s AGNTCY program, where libp2p is a partner. The initiative focuses on agentic, autonomous, and decentralized AI systems. Reliable cloud infrastructure is essential for hosting libp2p-based networking layers at scale.
Research and standards work presented at IETF 126 (Vienna) that a multi-region Akamai footprint would directly accelerate — global measurement vantage points, relay and bootstrap diversity, and edge nodes for agent-network experiments.
A common measurement framework for large-scale libp2p deployments: deterministic CBOR encoding, content-addressed telemetry, and cryptographic provenance that turns operational exhaust into reproducible scientific artifacts.
Early operational lessons from running AGNTCY over libp2p: quantifying the hidden discovery tax, relay dependency, and NAT complexity across cloud, enterprise, Kubernetes, edge, and residential environments.
Active participation in emerging agent-networking work — Discovery of Agents, Workloads and Named entities (DAWN), Dynamic Multi-agent Secured Collaboration (DMSC), and agent communication protocols — alongside Cisco, Ericsson, Huawei, Alibaba, and China Telecom.
The joint AGNTCY × libp2p stack: SLIM for multi-agent coordination, Kademlia DHT for capability discovery, and libp2p primitives (QUIC, WebRTC, WebSockets, TCP) as the transport-agnostic core with self-sovereign Ed25519 peer identities.
Cross-implementation conformance across Go, Rust, JavaScript, and Python nodes, with a unified testing harness producing comparable performance baselines and a Wireshark-style dissector layer for higher-level telemetry.
Research topics we would like Akamai's edge footprint to help answer: selective disclosure in verifiable telemetry, differential privacy for network observations, and taxonomy extensions for distributed inference.
$100B+
Secured across deployments
30+
Production networks
10+
Active implementations
300+
Monthly contributors
The capabilities we want to validate during the deep-dive.
Run the same deployment workflows on AWS, Akamai Cloud, and self-managed infrastructure with minimal changes.
Host retrieval-augmented generation and distributed AI experimentation infrastructure with predictable egress pricing.
Deploy reliable bootstrap, relay, and signaling nodes for peer-to-peer connectivity across regions.
Gain visibility into multi-cloud spend and usage as workloads grow 3–4x with user adoption.
Evaluate HIPAA-ready controls and hardened networking for future enterprise and healthcare-facing workloads.
Leverage Akamai’s distributed edge for content-addressed and globally distributed applications.
How we plan to use the $5,000 Akamai Cloud credits.
| Workstream | Primary Use | Expected Outcome |
|---|---|---|
| py-ipfs-lite | Compute and storage nodes for IPFS-lite testing, CI/CD build runners, and multi-region validation. | Validated deployment template and performance baseline on Akamai Cloud. |
| Universal Connectivity | Long-running bootstrap/relay nodes, NAT traversal test beds, and cross-provider peer simulations. | libp2p connectivity benchmark comparing Akamai, AWS, and self-hosted nodes. |
| Shared platform | Monitoring, logging, artifact storage, and documentation hosting for both projects. | Observable, documented multi-cloud reference stack. |
A high-level view of how Akamai Cloud fits into our stack.
Note: Detailed architecture diagrams and instance configurations will be shared ahead of the meeting to help the Akamai Solution Architect prepare a tailored recommendation.
Proposed 60-minute flow for the deep-dive.
Confirm attendees, agenda, and desired outcomes.
Walk through py-ipfs-lite, Universal Connectivity, and the Cisco AGNTCY alignment.
Verifiable telemetry (CBOR/VTO), agent-network measurement, and the standards roadmap.
Review AWS setup, spend and growth, then compute, storage, egress, edge, and observability on Akamai.
Define a 30-day proof of concept and map the $5,000 credit allocation.
Assign action items and schedule the follow-up session.
Actions we are preparing to discuss and confirm.
Share architecture diagrams and instance configurations.
Prepare a tailored architecture recommendation and credit usage plan.
Define a 30-day POC scope for py-ipfs-lite and Universal Connectivity.
Introduce US and Europe account representatives and grant/funding contacts.
Who will be on the call.
Manu Sheel
libp2p / Protocol Labs
Johanna Moran
Strategy Lead · libp2p
Luis Beltran
Account Executive · Akamai Cloud
Aastha Pabuwal
Cloud Consultant · Akamai
Praveena Reddy
Sr. Inside Sales Representative · Akamai
Key initiatives and communities tied to this meeting.